Prototype No. 2 – SWORD, Globus FTP and importing by reference

In parallel to our development using SWORD to try to solve the issue of uploading large research data¬† to Dspace, we’ve been working on another prototype which uses a bit of SWORD along with an FTP client (Globus) and Dspace’s own command line script for importing data into the archive.

The process works as so:

1) The user uploads their research data from their own device (PC, iPad, etc) up to ATMOS (remember we have this storage facility) using the Globus FTP client. We’ve played about with Globus and have initilly created a script to sort out the certificatation that Globus requires to transfer files between locations. Globus has many advantages here in that, in the event of any stoppage, the upload should be resurrected

2) Once uploaded the user uses a GUI that is build from the input-forms.xml file on the dspace server to enter metadata and choose the files on the server they have just uploaded. They can then submit to the archive.

Behind the scenes, Dspace is creating a set of files required for the command line ‘import’ script to submit the files by reference to the archive. In this way, the final stage works much quicker.

As you can see, this is a two stage process, with the user uploading the files and then submitting them to the repository, however, we have ideas to allow the user to do everything at the upload stage, or if the files sizes are not so large, there is a part to the GUI for simply uploading files straight to dspace as in a normal submission.

And where does SWORD come in? We use it to get the service document.

Testing continues…

 

Posted under Technical development

This post was written by Ian Wellaway on September 14, 2012

Comments are closed.

More Blog Post